Skip to content

Wire Briefings Search

Fetching from the wire…

01

02

03

04

05

06

07

08

Wire Briefings Search

Entity trail

SFSpeechRecognizer

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

1

Findings

1

Edges

2

Sources

1

Corpus findings

2026-07-14 / hn-researcherApple's New SpeechAnalyzer API Hits 2.12% WER, Beating Whisper Small and Its Own Legacy EngineA benchmark of Apple's new on-device SpeechAnalyzer API measured a 2.12% word error rate on clean English audio, versus 3.74% for Whisper Small and 9.02% for the legacy SFSpeechRecognizer, while running ~3x faster than Whisper Small on an M2 Pro. The author concluded it is now the strongest on-device English transcription option measurable on Apple hardware, though Whisper still wins on language coverage and cross-platform support. The post drew 546 points and 221 comments on HN.

Graph relationships

BENCHMARKED AGAINST
SpeechAnalyzer API
Apple's SpeechAnalyzer API achieved 2.12% WER versus 9.02% for legacy SFSpeechRecognizer.
Source finding
BENCHMARKED AGAINST
SpeechAnalyzer API -> SFSpeechRecognizer
Apple's SpeechAnalyzer API achieved 2.12% WER versus 9.02% for legacy SFSpeechRecognizer.
Source finding

Source trail

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues